Fast hog feature computation based on CUDA

Author

Chen Yan-ping ; Li Shao-zi ; Lin Xian-ming

Author_Institution

Cognitive Sci. Dept., Xiamen Univ., Xiamen, China

Volume

4

fYear

2011

fDate

10-12 June 2011

Firstpage

748

Lastpage

751

Abstract

Histogram of oriented gradients (HOG) is one of the most popular descriptors used for pedestrian detection, but this descriptor has its own drawback. Like most sliding window algorithms it is very slow, making it unsuitable for many real-time applications. This paper proposes a parallel implementation of the HOG algorithm. It bases on CUDA (compute unified device architecture) platform that could use parallel computing of graphic processing unit (GPU). The time consumption of HOG running on the GPU and on the CPU is compared by experiments in this paper. The results demonstrate that the HOG on GPU performs better than the HOG running on CPU, and is approximate 10 times speedup.

Keywords

computer graphic equipment; coprocessors; gradient methods; object detection; parallel processing; traffic engineering computing; CUDA; fast hog feature computation; graphic processing unit; histogram of oriented gradients; parallel computing; pedestrian detection; sliding window algorithms; CUDA; GPU; HOG; pedestrian detection;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Automation Engineering (CSAE), 2011 IEEE International Conference on

Conference_Location

Shanghai

Print_ISBN

978-1-4244-8727-1

Type

conf

DOI

10.1109/CSAE.2011.5952952

Filename

5952952